Improving Knowledge-aware Dialogue Generation via Knowledge Base Question Answering

TL;DR

This paper introduces TransDG, a novel knowledge-aware dialogue generation model that leverages techniques from knowledge base question answering to improve factual accuracy and relevance in open-domain dialogues.

Contribution

The paper proposes a new model that transfers question understanding and knowledge matching from KBQA to dialogue generation, with response guiding attention and multi-step decoding strategies.

Findings

TransDG outperforms existing methods on benchmark datasets.

The model generates more informative and fluent dialogues.

The approach effectively incorporates commonsense knowledge into open-domain conversations.

Abstract

Neural network models usually suffer from the challenge of incorporating commonsense knowledge into the open-domain dialogue systems. In this paper, we propose a novel knowledge-aware dialogue generation model (called TransDG), which transfers question representation and knowledge matching abilities from knowledge base question answering (KBQA) task to facilitate the utterance understanding and factual knowledge selection for dialogue generation. In addition, we propose a response guiding attention and a multi-step decoding strategy to steer our model to focus on relevant features for response generation. Experiments on two benchmark datasets demonstrate that our model has robust superiority over compared methods in generating informative and fluent dialogues. Our code is available at https://github.com/siat-nlp/TransDG.